Activity for ammonium formate and formic acid decomposition

Au/TiO₂, Au/La-TiO₂, Au/W-TiO₂ and Au/La-TiO₂-aged were tested for ammonium formate (AmFo) and formic acid decomposition in a simulated exhaust gas containing 10 vol% oxygen and 5 vol% water at different temperatures in the range 160 °C to 300 °C using various contact times (2.4*10^-5 g s.cm^-3 ≤ W/F ≤ 1 6*10^-4 g s.cm^-3) (Figure 5.4 left). These catalysts at identical gold loadings differ either by support modification at equal gold particle size (Au/TiO₂ versus Au/W-TiO2 versus Au/La-TiO2-aged) or by gold particle size on the same support (Au/La-TiO2

versus Au/La-TiO₂-aged). In agreement with the previous studies,^[25] the values for formic acid Au/La-TiO

2

Au/W-TiO

2

5 nm Au/TiO

2

Au/La-TiO

2-aged 5 nm

5 nm 5 nm

5 nm

Promotion of ammonium formate and formic acid decomposition on Au/TiO2 by support basicity

59

conversion obtained at 300 °C over all the catalysts were identical for formic acid and AmFo.

Ammonia was never oxidized.

Clearly, at temperatures ≥ 260 °C, the lanthanum-modified catalysts, both fresh and aged, exhibited higher formic acid conversion, while Au/W-TiO₂ showed the lowest performance.

Additionally, at these temperatures, the formic acid conversion obtained over the lanthanum-modified catalysts using formic acid as the precursor surpassed the corresponding values obtained over Au/TiO₂ using AmFo as the precursor. This observation implies that the promotional effect initiated by base modification surpasses the rate enhancement originating from the presence of a stoichiometric amount of ammonia in the feed gas.^[25] At 300 °C, when formic acid was used as the precursor, Au/La-TiO2 and Au/La-TiO2-aged exhibited close to 100% conversion at W/F= 7.5*10^-5 g s.cm^-3 and 1*10^-4 g s.cm^-3, respectively. In contrast, even at the highest contact time (W/F = 1.6*10^-4 g s.cm^-3), Au/TiO2 and Au/W-TiO2 did not completely decompose formic acid, resulting in 92% and 69% conversion, respectively. Using the lowest contact time (W/F = 2.4*10^-5 g s.cm^-3), the formic acid conversion obtained over the un-modified catalyst was <50% of that of Au/La-TiO₂ and more than twice that of Au/W-TiO₂. Under these conditions, the Au/La-TiO₂-aged still retained close to 60% of its original activity, thus, out-performing the fresh un-modified catalyst. At 260 °C and W/F ≥ 1*10^-4 g s.cm^-3, the formic acid conversion obtained over Au/La-TiO₂ using AmFo and formic acid as precursors reached close to 100% and any further increase in the contact time only led to increased carbon dioxide selectivity. Au/La-TiO₂-aged suffered a decrease in conversion owing to the thermal treatment, reaching ~80% conversion at the highest contact time, still exceeding the performance of Au/TiO₂ by more than 15%. At 200 °C, when AmFo was used as the precursor, the formic acid conversion ranged from 70% over Au/La-TiO2 to 50% and 28% over Au/TiO2 and Au/W-TiO2, respectively.

At lower temperatures (≤ 200 °C), the difference in the activities of the un-modified and the base-modified catalysts decreased when formic acid was used as the precursor and even turned in favor of the un-modified catalyst, when AmFo was used as the precursor. This has two implications: (i) base modification does not improve the formic acid decomposition activity at lower temperatures, and (ii) the promotional effects derived from the gas-phase and the base modification of the catalyst operate in complementary temperature regimes, wherein, at lower temperatures (≤ 200 °C), the ammonia-induced rate enhancement ^[25] dominates. While, the interaction of ammonia is more favored at lower temperatures,^[25] the catalytic effect prevails predominantly at higher temperatures, which reconciles with the higher E_a values stemming from the highly negative formic acid orders observed over the base-modified catalysts (See section 3.3). Under all the tested conditions, the acid-modified catalyst was the worst performing catalyst.

0

Figure 5.4 Formic acid conversion as a function of contact time, expressed as W/F (g s cm^-3), obtained using AmFo (closed symbols) and formic acid (open symbols) at different temperatures (left) and temperature dependence of formic acid conversion (closed symbols) and carbon dioxide selectivity (open symbols) at W/F = 2.4*10^-5 g s cm^-3 (right) over Au/TiO2, Au/W-TiO2, Au/La-TiO2 and Au/La-TiO2-aged catalysts.

Promotion of ammonium formate and formic acid decomposition on Au/TiO2 by support basicity

61 4.0x10^-5 8.0x10^-5 1.2x10^-4 1.6x10^-4 0

4.0x10^-5 8.0x10^-5 1.2x10^-4 1.6x10^-4 260 C

W/F, g.s.cm^-3

Figure 5.4 (right) depicts the temperature dependence of formic acid conversion and carbon dioxide selectivity over Au/TiO₂, Au/La-TiO₂, Au/W-TiO₂ and Au/La-TiO₂-aged at W/F = 2.4*10^-5 g s.cm^-3. A progressive drop in carbon dioxide selectivity was evidenced over all the catalysts above 200 °C, with carbon monoxide being the only other (undesired) product of formic acid decomposition. Evidently, the low activity of the acid-modified catalyst is accompanied by the lowest carbon dioxide selectivity compared to the un- or base-modified catalysts. The thermal aging at 600 °C did not diminish the carbon dioxide selectivity, which remained >98% over the entire temperature window. At 300 °C, Au/TiO2 exhibited 81% carbon dioxide selectivity, while Au/W-TiO₂ only 70%. At temperatures ≤ 200 °C, there was no carbon monoxide production over any of the catalysts. Figure 5.5 (top) compares the effect of support modification on the carbon dioxide/carbon monoxide ratio at 300 °C and 260 °C at different contact times. The carbon dioxide/carbon monoxide ratios observed with lanthanum-modification remained practically unaffected with aging.

Figure 5.5 Evolution of CO₂/CO ratio as a function of contact time obtained from formic acid decomposition over Au/TiO2, Au/W-TiO2, Au/La-TiO2 and Au/La-TiO2-aged catalysts at 300 °C and 260 °C ( top) and formic acid conversion versus carbon dioxide selectivity obtained over Au/TiO₂, Au/W-TiO₂, Au/La-TiO₂ and Au/La-TiO₂-aged catalysts at 300 °C (bottom).

0 20 40 60 80 100

There was a striking enrichment of carbon dioxide in the product gas, while the carbon monoxide formation remained suppressed over the base-modified catalysts. At the two temperatures and W/F = 2.4*10^-5 g s.cm^-3, the carbon dioxide/carbon monoxide ratio experienced a ~10-fold increase upon base-modification, while acid-modification decreased the ratio by 50%. Hence, base-modification of the catalyst selectively promotes formic acid decomposition to carbon dioxide, reminiscent of the effect of ammonia.^[25] Figure 5.5 (bottom) compares the carbon dioxide selectivity versus formic acid conversion obtained using formic acid at 300 °C over Au/TiO2, Au/La-TiO2, Au/W-TiO2 and Au/La-TiO2-aged. The base- and the acid-modified catalysts exhibited the highest and the lowest carbon dioxide selectivity, respectively, over the whole conversion range. At all contact times, the carbon dioxide selectivity was >98% for the base-modified catalysts, while it dropped to 80% and 70% over Au/TiO₂ and Au/W-TiO2, respectively, at W/F = 2.4*10^-5 g s cm^-3. The shaded region shows that at similar conversion levels over the four catalysts, a marked difference exists in the carbon dioxide selectivity. This strongly suggests that base-modification increases the intrinsic propensity to produce carbon dioxide from formic acid decomposition.

Figure 5.6 Oxidation of 650 ppm carbon monoxide (gas-phase) over Au/La-TiO₂ and Au/La-TiO₂-aged in the presence (closed symbols) and absence of water (open symbols) (left) and comparison of carbon dioxide produced from formic acid decomposition (bars) and gas-phase carbon monoxide oxidation (scatters) as a function of temperature over Au/TiO₂, Au/W-TiO₂, Au/La-TiO₂ and Au/La-TiO₂-aged catalysts using 650 ppm formic acid and 650 ppm carbon monoxide at W/F =1.6*10^-4 g s cm^-3 (right).

Instead of direct formic acid oxidation to carbon dioxide, formic acid could alternatively first decompose to carbon monoxide as an intermediate, which subsequently oxidizes to carbon dioxide.^[192,193] Gas-phase carbon monoxide oxidation experiments were performed to assess the contribution of this reaction pathway towards increased carbon dioxide production from formic acid decomposition over the base-modified catalysts. Figure 5.6 (left) plots carbon monoxide conversion as a function of temperature over the fresh and aged base-modified

Promotion of ammonium formate and formic acid decomposition on Au/TiO2 by support basicity

63

water adversely affected carbon monoxide conversion. Under realistic conditions involving 10 vol% water, the carbon monoxide conversion over both the fresh and aged base-modified catalysts was 50% of those when no water was present. Haruta et al. reported that the influence of water shifts from being positive to detrimental with increasing concentration.^[194] They found that above 200 ppm, water suppressed the carbon monoxide oxidation activity of Au/TiO₂. There was a conspicuous decline in carbon monoxide conversion with aging, which is expected to result from the sintering of gold during the high temperature treatment.^[195,196]

Figure 5.6 (right) hypothetically compares the carbon dioxide production from decomposition of formic acid with the oxidation of isoconcentrations of carbon monoxide (650 ppm each) over the four catalysts under identical feed conditions (W/F = 1.6*10^-4 g s.cm^-3). At 300 °C, close to 100%

carbon dioxide yield was obtained from formic acid decomposition over both the fresh and aged base-modified catalysts, while the carbon monoxide conversion under identical conditions was only 18% and 10%, respectively. The un-modified catalyst exhibited 80% carbon dioxide yield from formic acid, while converting 11% carbon monoxide. The acid-modified catalyst produced 54% carbon dioxide from formic acid decomposition and showed close to negligible conversion for carbon monoxide oxidation. At 160 °C, there was no carbon monoxide conversion over any of the catalysts. At the same temperature, Au/TiO₂, Au/La-TiO₂ and Au/La-TiO₂-aged produced 17%, 19% and 13% carbon dioxide, respectively, while Au/W-TiO₂ yielded 10% carbon dioxide.

Since, the carbon monoxide conversion is significantly lower or negligible compared to the total carbon dioxide produced from formic acid decomposition over all the catalysts under the investigated conditions, there must exist an independent pathway for carbon dioxide formation directly from formic acid that precludes carbon monoxide formation and its subsequent oxidation. Therefore, it can be ascertained that base modification selectively promotes this direct formic acid conversion pathway responsible for carbon dioxide formation.

Figure 5.7 illustrates the relative percentage increase in the pseudo first order mass-based rate constants for formic acid decomposition upon introduction of a stoichiometric amount of ammonia as a function of temperature over Au/TiO2, Au/La-TiO2, Au/W-TiO2 and Au/La-TiO2 -aged. Au/TiO₂ exhibited the most pronounced effect of ammonia, followed by the base- and acid-modified catalysts. Aging did not alter the effect of ammonia. At 160 °C, in the presence of one molar equivalent of ammonia, the formic acid decomposition rate underwent close to 190%

increase over the un-modified catalyst, while the base- and acid-modified catalysts showed

~80% and 45% increase, respectively. At 200 °C, the extent of enhancement reduced to 74%,

~23% and 21% over the un-, base-, and acid-modified catalysts, respectively. At 300 °C, the effect of ammonia on formic acid decomposition was negligible over the three catalysts.

Dumesic and co-workers found that modification of alumina by lanthanum caused a decrease in the initial heat of ammonia adsorption by reducing the number of acid sites and simultaneously

180 240 300 0

50 100 150

200 Au/TiO₂

Au/La-TiO₂ Au/W-TiO₂ Au/La-TiO₂-aged

Increase in rate constants, %

Temperature,C

increasing the number of basic sites whose strength increased with increasing loading of lanthana.^[197] The lower extent of ammonia-induced-increase in rates over the base-modified catalysts can be rationally attributed to a lower level of interaction between ammonia and the basic surface of the catalysts. Even though the trend associated with the acid-modified catalyst appears counter-intuitive at first sight, it can be easily explained by taking into account the acidity of Au/W-TiO₂, that is likely to result in very high ammonia coverages.^[198] This in turn can be speculated to be detrimental causing unfavorable competition with formic acid, leading to the lowest degree of enhancement compared to un- and base-modified catalysts.

Figure 5.7 Effect of 1 molar equivalent of ammonia on formic acid decomposition rate constants over Au/TiO₂, Au/W-TiO₂, Au/La-TiO₂ and Au/La-TiO₂-aged at W/F = 2.4 * 10^-5 g s cm^-3.